Genetic risk score calculated using SNPs with mixed effect directions

Question

[krishnapashu912]

Hello,

I was wondering about the following question,

SNPs used for calculation of genetic risk score (GRS) can have effect sizes in both directions (+ or -). Isn't it important to use SNPs with effect size in same direction? Wouldn't we cancel out the effect of risk alleles if we mix SNPs with opposite directions?

I would appreciate for any insights!

Answer 1

[Sam]

Assuming the + effect size = allele associated with increased risk of disease, and - effect size = allele associated with decreased risk of disease. If the effect sizes have the same magnitude, and when you have the same amount of "risk" alleles and "protective" allele, then you can argue that the protective effect "cancel" out the risk. By this logic, you can have people who have a negative risk score, and in this mock example, they are "protected" from having the genetic disease